The LiFePO4 Headway cell 38120P TEST REPORT inside

[OneEye]

I believe the ~$17/cell was an at-factory quote on 1,000 minimum order. Extremely doubtful we will see a price like that for delivered small quantities.

 

[Doctorbass]

Here is the cell no 3 discharge graphs test report at 0.125C, 1C 2C 3C and 4 C:
they are supposed to hold 10C continuous and 20C for short period... I will test that soon

I'm pretty impressed by the low Peukert effet about capacity that is not very affected at different discharge currents.

they also stay under 31 degree C and they charge pretty fast at 10A. (1.2C)

Please note that my CBA analyzer is not perfect (...from now.. :wink: ) and that at 1A the voltage drop between the direct battery tab and the volt reading is 5mV so in the graph you should read 5mV higher than the volt value (offset error) and that at 32A this offset is 170mV. that mean that as the current is high as the error increase and that at 32A the mean voltage should be 2.95V instead of 2.78V

I can conclude that with this cell no 3, the energy in Wh should be around 23Wh at 32A and around 27.5Wh at 1A

the results are prety similar to the A123 discharge graph at different C rate

coming next; cell no 1 and 2 tests

Doc

 

[drewjet]

Doc,

Those discharge graphs look good, not as good as A123, but then who can match them.

I would love to see long term 1000 cycle test results, at full discharges.

Hoping to hear prices soon!

 

[wanders]

Those discharge charts actually make the Headway cells look better than A123 to me. Sure, the "cliff" isn't quite a sharp, but if you compare the relative discharge rates, the overall voltage sag of the Headway cells is significantly less for higher rates of discharge (if I am reading the charts correctly).

Willie

 

[safe]

Urgent Data Request

The group as a whole likes to see the charts of the cell performance, but I'm urgently requesting that you post the data for the charts so that folks like me can do analysis with them.

What I'm really looking for is a constant "C" rate discharge for each cell... that's one set of data per cell per discharge... so that I can compare how they differ between themselves. From this data I can evaluate the effectiveness of balancing theories.

If you can post the data as simple "Text & Commas" (the "*.csv" extension) then I don't have to worry about Excel format related problems...

 

[disadvantage]

(1000 cell minimum order) / (16 cells per pack) = 62 battery packs, using the new G&B BMS.

Half as many if you are building doubled 16 AHr packs.

Still, that's $17,000 worth of Chinese lithium, and shipping is extra.

 

[docnjoj]

Leaves us with about 300-350 plus cms per pack. Not that bad for good quality cells! They seem to have decent solder tabs for those of us who solder packs together.
David

 

[Doctorbass]

Leaves us with about 300-350 plus cms per pack. Not that bad for good quality cells! They seem to have decent solder tabs for those of us who solder packs together.
David

Here is the (-) tab spotwelded

It seems suffusent to hold the claimed 160A briefly and 80A continuously.

here is a pic i took few minutes ago:

 

[Doctorbass]

i'm cycling cell no 1 for 10 complete cycle at 10A charge and 5A discharge that is the maximum that my RC tester can do.

That should take around 22hour. I have 3 complete cycle done from now.

I found something interesting in the Headway site see pics below:

A 12V 20A on a ...Hyunday using 8 of these cells.

Also, an interesting auction here 48V 20A made of 16 of these cells 100A continuous 200 max disch.. i'm wondering what is their BMS?
http://www.austinev.org/evtradinpos...ertisement_active&fromfrommethod=showhtmllist

Doc

 

[Doctorbass]

Safe, your graph data will come soon!

For now here are the first 10 cycles of the cell 1 at 5A disch cut 2.9V 10A charg 10 min pause between cycles

looking at the graph below, i can conclude that the cell no 1 lost 80mAh during the first 10 cycles. That's a bit less than 1% of its total capacity.

By taking account that li-ion cells capacity is based on how many cycles it can reach before droping to 80% of the original capacity, that would mean that 1% loss for 10 cycles= 200cycles for 20% drop capacity.... BUT.. that cell is brand nre and need to stabilyze and we can't make conclusion at this time with only 13 or 14 cycles total.

Doc

 

[Jeremy Harris]

Following Doc's link to the AustinEV site above, I got a quote from Victoria at Headway today. Very prompt response, seems pretty professional to me.

The price for 48off 38120 10Ah cells is £9.56 each, plus shipping (about $19 US each).

It looks like they are happy to sell small quantities at a reasonable price, which is good news.

I look forward to the next instalment of your tests, Doc!

Jeremy

 

[safe]

Safe, your graph data will come soon!

Don't forget to save the data as either "text" or "csv". That way it's not bound to any particular version of a spreadsheet.

I would think that if you simply laid out the data from left to right and made each cell be equal to some unit of time then it would be easy to process. But if the data comes out differently I'll do whatever I can to convert it.

 

[Toorbough ULL-Zeveigh]

I'm somewhat confused, a chronic state, only moreso by this comment that implies these cells are supposedly *not* from China?

http://www.bikeforums.net/showthread.php?t=370920

 

[wanders]

For now here are the first 10 cycles of the cell 1 at 5A disch cut 2.9V 10A charg 10 min pause between cycles
...
By taking account that li-ion cells capacity is based on how many cycles it can reach before (dropping) to 80% of the original capacity, that would mean that 1% loss for 10 cycles= 200cycles for 20% drop capacity.... BUT.. that cell is brand (new) and need to (stabilize) and we can't make conclusion at this time with only 13 or 14 cycles total.

Great work, Doc - I love the scientific approach

Your chart implies charge cutoff at ~3.75 V and discharge cutoff at 3.2 V, while your text states a discharge cutoff at 2.9V. If you were truly using the chart voltage markers, the capacity given would be excellent. Could you elaborate?

Also, since the cells are rated at 8AH, you are still working in the excess capacity regime. As you know, due to the nature of the electrochemistry of the electrodes in a cell, there is some "normalizing" that takes place when the cells are new. As you have said, it is a little early to estimate the cycle lifetime of these cells.

Thanks again for the excellent work!

Willie

 

[Doctorbass]

For now here are the first 10 cycles of the cell 1 at 5A disch cut 2.9V 10A charg 10 min pause between cycles
...
By taking account that li-ion cells capacity is based on how many cycles it can reach before (dropping) to 80% of the original capacity, that would mean that 1% loss for 10 cycles= 200cycles for 20% drop capacity.... BUT.. that cell is brand (new) and need to (stabilize) and we can't make conclusion at this time with only 13 or 14 cycles total.

Great work, Doc - I love the scientific approach

Your chart implies charge cutoff at ~3.75 V and discharge cutoff at 3.2 V, while your text states a discharge cutoff at 2.9V. If you were truly using the chart voltage markers, the capacity given would be excellent. Could you elaborate?

Also, since the cells are rated at 8AH, you are still working in the excess capacity regime. As you know, due to the nature of the electrochemistry of the electrodes in a cell, there is some "normalizing" that takes place when the cells are new. As you have said, it is a little early to estimate the cycle lifetime of these cells.

Thanks again for the excellent work!

Willie

the cut off IS at 2.9V... the 3.2V you see is after 10 minutes no load... at the end of the pause time between cycles

 

[Jeremy Harris]

Doc,

Do you think that these cells will be OK for high rate of discharge operation?

I'm thinking of a 16 series, 3 parallel, pack, 48V @ 30Ah with max current of around 120 to 150 amps, continuous current of around 50 amps or so.

Jeremy

 

[Doctorbass]

I just finished the lasts 10 cycles for a total from now of 20 complete cycles. here are the data:




Doc

 

[Doctorbass]

Doc,

Do you think that these cells will be OK for high rate of discharge operation?

I'm thinking of a 16 series, 3 parallel, pack, 48V @ 30Ah with max current of around 120 to 150 amps, continuous current of around 50 amps or so.

Jeremy

For sure these cell can handle 50A continuous!

at 30A continuous, they reach 39 degree C frtom the ambient of 24. I expect they should reach around 50-55 degree C at 50A. In a scealed bex maybe they could reach around 60 degree C .

their internal resistor is around 7milliohm so at 50A each cell will dissipate around : P= R x I^2
= 17.5W each. at 50A these cell give around 7.5Ah. 50A would mean each cells would give that power from full to empty in around 7.5Ah/50A = around 9 minutes so that represent a loss of Wh in heat of 17.5W x(7.5Ah/50A)= 2.62Wh of energy loss per cell at 50A. Each cell have (7.5Ah x 2.7V)= 20.25Wh at 50A and loose 2.62Wh in heat for a total Wh of 22.87Wh of energy out of the cell.


Doc

 

[MyPC8MyBrain]

MMMMMMMMM tasty.
Wonder how much these are going for....

48V 20AH Hedway packs with BMS

http://headway-cn.en.alibaba.com/pr...48v20Ah_LiFePO4_lithium_ion_battery_pack.html

 

[safe]

I just finished the lasts 10 cycles for a total from now of 20 complete cycles. here are the data:

I'm glad you posted your results, but I was looking for more than one cell. The idea was to spot the differences between cells and log that data so that I could do some balancing analysis. What I'm really looking for is the moment-to-moment drain performance for each cell in a single cycle... but multiple cells...

 

[Doctorbass]

I just finished the lasts 10 cycles for a total from now of 20 complete cycles. here are the data:

I'm glad you posted your results, but I was looking for more than one cell. The idea was to spot the differences between cells and log that data so that I could do some balancing analysis. What I'm really looking for is the moment-to-moment drain performance for each cell in a single cycle... but multiple cells...

Safe,

An accurate cell testing need more than a a single set of 10 cycles..allmost if we only talk about the 10 firsts cycles of that cell.

it take around 55 minutes to charge and 1h30 to discharge + 10 minutes pause between each cycles.

55+90+10=155min = 2h35 per cycles. a complete 10 cycles take a day to be performed. I have 3 cells and they need at least 50 cycles. I have ONE tester.. doing some math we get more than 2 weeks NON STOP!.. But i have a job and i'm not always close to my tester when each cycles serie are finished to se the next one....
5A discharge and 10A charge is the fastest rate i can get and also a good trade off to not dammage the cell and get a good efficiency about the discharge and the charge process.

From now, i'm completing the third 10 cycles serie. you have already the first 20 cycles of the first cell + thedischarge curve of cell.

Be patient Safe!


Doc

 

[safe]

Be patient Safe!

I will take what I can get...

 

[oatnet]

I just finished the lasts 10 cycles for a total from now of 20 complete cycles. here are the data:

Wow Doc, great work! I have been thinking about these cells a lot and it is interesting to see the numbers, thanks for sharing your information.

Your 20-cycle test shows pretty severe, and linear, degradation of capacity at 100% DOD. How does that compare with other LiFe you have tested?

I wonder if they will fare better at 70% DOD (my preferred discharge cutoff) in real-world usage - but I don't see a good way to test for that in the lab.

-JD

 

[Doctorbass]

Cell 1 First 30 cycles completed

It seems that the temp and the great temperatures outdoor influenced the cycling of these 21-30 cycles..

The capacity increased with only 3-4 degree celcius rise.. I can not stabilyze perfectly the temp in my lab.

 

[Jeremy Harris]

That's great work you're doing, Doc. It's good to know that it looks like the gradual drop in capacity you measured was a function of cell temperature, rather than a sign of the cells degrading.

I'm on the verge of buying a stack of these cells for my bike. Do they come with tabs as standard?

I'm thinking of building either a 16S, 3P or perhaps a 16S, 4P pack, that needs to be able to deliver over 100 amp peaks OK.

Jeremy